The overall goal of this proposed SBIR project is to introduce a powerful new non-linear estimation framework for three-dimensional structure refinement of protein molecules using X-ray crystallographic data. This framework will be enabled by ASTER-X (Automated Structure Estimation and Refinement for X-ray); a new code that promises to increase the throughput of crystallographic structures via improved speed and larger radius of convergence. This increased throughput is essential to link structural biology to the exploding fields of genomics and proteomics. The specific aims of the proposed Phase I work are twofold. The first goal is to implement a powerful refinement method for the early crystallographic model-building/refinement stages that will reduce the amount of human intervention and expertise necessary for this process. We will evaluate the power of ASTER-X in terms of radius of convergence and speed using a relatively straightforward, scalable, molten box approach to traversing the structure; an approach that can be implemented in an interactive mode. The second goal is to develop a more advanced version of ASTER-X that will allow treatment of the entire crystal structure at one time, and will thus enable reciprocal space refinement. This advanced version promises speedups of one to two orders of magnitude over state-of- the-art refinement methods. PROPOSED COMMERCIAL APPLICATIONS: The objective of this SBIR project is to bring to market an innovative new software product (ASTER-X) for protein 3D structure determination from X-ray crystallographic data. ASTER-X offers increased throughput of crystallographic structures via improved speed and larger radius of convergence. ASTER-X will be highly desired by researchers in pharmaceutical and biotechnology industries, as well as by those in academia, especially as it provides a link between structural biology and the exploding fields of genomics and proteomics.